A purchase order is not collateral. It is, however, the closest thing a manufacturer has to evidence about the future, and every capacity request is a forecasting exercise whether the paperwork admits it or not. When a shop asks for money to add a cell, open a second shift or stand up a line, the desk is not really evaluating the machine. It is evaluating three things in sequence: whether the work exists, whether the buyer will pay for it, and whether this shop can convert raw material and labor hours into shipped parts at the margin it claims.
That is a different conversation from a generic working-capital request, and manufacturers who understand the difference get better structures out of it. Capacity capital sized against a documented order book can be amortized against the revenue that capacity produces, which is a coherent story an analyst can write up. Capacity capital sized against optimism gets treated as an unsecured bet on a small business and priced or declined accordingly. This guide covers what a credit desk does with a backlog report, how capacity requests get sized, and the gap between installing equipment and getting paid for what it makes.
What an order book proves, and what it does not
An order book proves demand. It does not prove profit, and it does not prove capability. Demand is documented, so it is the easy one. Profit is harder: shops win volume by quoting thin and then find the new work funds less debt service than the old work did. Capability is the one operators underweight, a machine does not add capacity if the real constraint is a heat-treat vendor, a single programmer, or an inspection department already running behind.
So the first thing to do with a backlog is sort it by how binding it actually is. Not all paper is the same paper.
| Document | What it commits the buyer to | Typical weight in sizing |
|---|---|---|
| Firm PO with quantity, price and delivery dates | A specific quantity at a specific price | Strongest. Can be sized against directly, subject to buyer credit |
| Blanket order with scheduled releases | A ceiling, not a floor: releases drive real volume | Strong where release history exists; weighted to trailing releases |
| Long-term supply agreement, no minimums | A position on the bid list and agreed pricing | Useful context, rarely sized against on its own |
| Award letter or letter of intent | Intent, usually conditional | Directional. Carries more when the program and start date are named |
| Customer forecast or planning schedule | Nothing contractual | Read as planning data and discounted heavily |
Those weightings are illustrative and vary by lender, program and file. The pattern behind them does not vary much: the more specific the commitment and the better the buyer's credit, the more of the backlog survives into the sizing math.
Sizing the request against contribution, not revenue
Operators present capacity requests in revenue terms. This cell brings in eight hundred thousand a year. A desk converts that to contribution margin almost immediately, because revenue does not service debt. Contribution does.
Contribution margin on a production job is the sale price less the variable cost to make it: material, direct labor including the burden that actually varies, outside processing, perishable tooling, freight and scrap allowance. What is left is what funds the machine payment, the incremental overhead the expansion creates, and the profit. If the incremental contribution does not comfortably cover the incremental debt service with room left over, the expansion is not self-funding and the rest of the business has to carry it, which is a much harder file.
Run that same arithmetic honestly before you ask. If the second calculation lands near 1.10x, the request is too large for the order book supporting it, and the fix is either a smaller machine, a longer amortization, a deposit from the customer, or a price conversation with the buyer, not a better presentation.
Whose credit is really behind the order
When a request leans on one customer's volume, that customer becomes part of the credit. A large buyer with public financials strengthens the file. A privately held tier-one with a thin backlog of its own does less, and a startup issuing a large first order can weaken it, because the desk now has to underwrite two businesses instead of one.
Payment terms matter as much as buyer quality. An order at net 90 consumes far more working capital than the same order at net 30, and that consumption is a separate financing need arriving before the first invoice is paid. Manufacturers regularly finance the machine and forget to finance the receivable the machine creates.
Customer concentration and how to present it
Concentration is the standard objection in job-shop and contract-manufacturing files, and pretending it is not there does not work, the desk will build the concentration table from your sales ledger anyway. Present it first, with the context that softens it.
The context that genuinely helps: relationship length in years rather than orders, the number of distinct part numbers inside that customer, whether you hold sole-source tooling, switching cost in qualification terms, and whether the relationship survived a downturn. A shop at 55 percent concentration with fourteen years of history and a qualified, tooled-up position is a different risk from a shop at 55 percent because one buyer placed one large order last year.
The ramp gap nobody budgets
The interval between signing for a machine and being paid for parts it produced is longer than almost every capacity model assumes, and it is a cash-consuming interval from end to end. This is the single most common reason a well-underwritten expansion still strains the business.
- Deposit at order. Frequently a meaningful share of the purchase price, paid before anything ships.
- Build and lead time. Varies widely by builder, machine class and configuration: plan it, do not assume it.
- Rigging, foundation, power and air. Site work often falls outside the equipment financing entirely.
- Install, runoff and operator training. Days to weeks of paying for a machine that is not producing revenue.
- First article, qualification and PPAP where required. Weeks more before saleable production is authorized.
- Production ramp to rate. Scrap runs high and cycle times run long until the process is dialed in.
- Invoice and collection. Then the customer's terms, which is where the receivable sits.
Added up, the working-capital requirement in front of first collection is often larger than the down payment on the equipment. Financing the machine and nothing else is how a profitable expansion produces a liquidity problem in month four.
Structures that fit a capacity request
Equipment finance against the asset
The default when the request is a machine with resale value. Terms generally track useful life, the asset secures itself, and the paper does not consume a bank line you may need elsewhere.
Term debt against the business
Better when the expansion is a package (machine plus tooling plus site work plus hiring) because proceeds are not confined to invoiceable equipment. Underwritten on coverage rather than collateral value.
A revolver behind the order book
The piece most often missed. Sized against receivables and eligible inventory, it funds the material and the terms gap new volume creates. Not an alternative to equipment financing: the other half of it.
Customer participation
Tooling payments, progress payments on long-lead work, or a deposit against a program launch. Cheapest money in the stack when the buyer will do it, and worth asking before you finance around it.
What to put in front of the desk
The capacity request packet
- Backlog report by customer and part number, showing quantity, price, delivery dates and document type for each line.
- Copies of the binding documents themselves: POs, blanket orders with release history, award letters.
- Trailing twelve months of release history for any blanket order you are sizing against.
- Costed quote or routing for the new work: material, cycle time, labor rate, outside processing, scrap allowance.
- Current capacity utilization by work center, showing where the constraint actually is today.
- Vendor quote for the equipment, plus separate quotes for rigging, foundation, electrical and tooling.
- A month-by-month ramp schedule from deposit through first collection, with the cash requirement at each step.
- Customer concentration table with relationship length, program count and qualification status.
- Two years of business tax returns, year-to-date P&L and balance sheet, and a current debt schedule.
Can I get financing on the strength of a purchase order alone?
Rarely on the PO alone. A firm order from a creditworthy buyer strengthens a file, but most structures still underwrite the manufacturer's ability to perform and its existing cash flow. Order-linked facilities exist and tend to be narrower and costlier than term structures, so they suit a one-off surge better than a permanent capacity change.
How much backlog is enough to support an expansion?
There is no fixed multiple. A desk is testing whether documented, binding volume produces contribution that covers the new debt service with real headroom after the overhead the expansion adds. Incremental coverage and whole-company coverage both get looked at, and thresholds vary by lender and program.
My largest customer is 60 percent of revenue. Is that automatically a decline?
No, but it must be addressed head-on. Relationship length, multiple part numbers inside the account, tooling you hold and qualification barriers all reduce the practical risk of a sudden loss. Expect concentration to affect structure, covenants or guarantor requirements rather than to end the conversation outright.
Should I finance the tooling and fixtures with the machine?
Where the structure allows it, bundling usually beats paying cash for soft costs and then running short during the ramp. Tooling has weaker collateral value, so some structures cap how much can be included. Ask early: rebuilding the request after a term sheet is issued costs time.
What if the new work is quoted at a lower margin than my existing mix?
Say so and show the reasoning. Volume work at a lower unit margin can still be a good credit if it absorbs fixed overhead you already pay. What damages a file is a costed quote that quietly assumes best-case cycle times and zero scrap, because the actuals surface in next year's financials.
How do I handle a long machine lead time when the order starts sooner?
Two common paths: subcontract or rent capacity to hold the delivery commitment while the machine is built, or negotiate a staged launch with the customer. Both cost margin in the interim, and both belong in the ramp schedule so the financing covers the bridge rather than assuming revenue starts at install.
Where to start
Sort your backlog into the five document categories above and total only the top two. That number, converted to contribution margin and reduced by the overhead the expansion adds, is what you can realistically ask to amortize against. Then build the ramp schedule from deposit to first collection and find the deepest cash point.
If those two figures both work, you have a file. If the second one is the problem, the request is not a bigger machine: it is a revolving facility behind the order book, and that is a different and usually easier conversation to have first.
Qualified Commercial Underwriting Desk
Credit and capital markets
The Qualified Commercial underwriting desk reviews commercial real estate, dealer and Main Street files daily. The Academy is written from that work (how files are actually read, priced and declined) rather than from a rate sheet.
Educational content only. Nothing here is a commitment to lend, an offer of credit, or tax, legal or accounting advice. Program terms, timelines and thresholds vary by lender, file and market conditions, and any figures shown are illustrative.